MARSHITE AND OTHER MINERALS FROM , CHILE

O. W. J.tnnnrl,, Haraard. Unitersity, Combrid,ge,Mass. A detailed account of the mineralogy and paragenesisof the deposit at Chuquicamata, in the arid Atacama Desert of north Chile, has recently been published by M. C. Bandy (1938). While Dr. Bandy was carrying out his investigation at Harvard University, the present writer was employed as a staff geologist at the mine. During that time and since the publication of Bandy's paper, several additional minerals have been found at Chuquicamata. Of these, salesite, a new mineral, has recently been described (Palache and Jarrell, 1939), and the present paper is intended to list the others, and thus make as complete as possible the description of the mineralogy of this remarkable deposit.

M,qnsnnn CuI Marshite, the natural cuprous iodide, was discoveredby C. W. Marsh (1S93)in the oxidized zone of the lead, zinc, and deposit at Broken HilI in the arid region of western New South Wales. Spencer and Prior (1901) published a complete description of the crystals from Broken HilI. Aminofr. (1922) has investigated the , and various other workers have prepared or studied synthetic material. In 1937, some crystals intimately associated with atacamite were found at the south end of the oxidized ore body at Chuquicamata. Be- cause of their color, isotropic nature, and high index of refraction, they were suspected of being marshite. Dr. Berman later confirmed this identification. Recently Mr. Lester Zeihen of the mine stafi at Chuqui- camata has sent additional specimensto Harvard University. CrystallogropZy. Marshite is isometric hextetrahedral. Professor Pa- lache examined the crystals and found that the principal faces present are the cube (100) and tetrahedron (111), which, due to oscillation, give the crystals a striated appearance, as illustrated in Fig. 1. The zone between the cube and tetrahedron is rounded due to the efiect of tris- tetrahedral faces. Several crystals were measured, one of which yielded definite signals from the two forms m(Il3) and B(223), shown in Fig. 2. The striated zonesgave weak reflectionsfrom other forms, none of which however is assured. Spencer (1901) has commented upon the similarity of the crystallography to that of sphalerite, and Aminoft (1922) has shown the crystal structure of the two minerals to be comparable. 629 O fiT. .TARRELL

Physical' properties. The crystals, which range up to several milli- meters across, are transparent and have an adamantine luster. When first found, they are colorlessor light honey-colored. Some of them were kept for about a year at Chuquicamata, and showed no sign of change. Recently, however, some of the crystals that had been kept in darkness at Cambridge, but not protected from the air, have become salmon to reddish colored, although otherwise remaining intact. At the same time other crystals, kept in glass-stoppered bottles, but exposed to light, have remainedcolorless;hence the changeto a reddish color seems.tobe promoted by contact with humid air, rather than by exposure to light.*

Frc. 1. Marshite. Frc. 2. Marshite.

The of the Chuquicamata material is only faint yellow as contrastedwith the bright yellow streak that Spencer(1901) considered so striking. Mr. Zeihen fi.rst reported that marshite was fluorescent. The color under the ultraviolet light is dark red. The average of several determinations of the density on the micro- balance gave a value of 5.68. The density calculated from Aminofi's dimensions for the unit cell is 5.60. A small prism was cut and polished on one of the crystals, and the indices of refraction, shown in Fig. 3, for various wave lengths were determined on the one-circle goniometer by the minimum deviation method with a monochromatic illuminator. The indices are slightly lower than those obtained by Spenceron the Broken Hill material, but this small discrepancy is perhaps explained by the slight difierence in chemical composition of the marshite from the two localities. The ex- ceedingly high dispersion, exceeding that of diamond, is noteworthy.

* It has been suggested that the reddish discoloration may be due to a chemical reac- tion with the enclosing paper. This has not, as yet, been verified. MARSHITE FROM CHILE 631

INDEX OF REFRACTION OF MARSH IT E FROM CHUQUICAMATA,CHILE

c

o

t c

Wave length

Frc. 3

Chemical co?nplsition. The analyses of the Broken Hill arrd Chuqui- camata material, together with the ideal composition of CuI, are given below.

Coupostrtor ol Mansrttr

tzJ Cu 33.0170 32.35 33.37 r 66.67 65.8.5 66.63 Ag None I.l9 cl 0.33

100.01 99.39 100.00 1. Marshite from Chuquicamata, Chile. F. A. Gonyer, analyst. 2. Marshite from Broken Hill, N.S.W. G. T. Prior, analyst (1902). 3. Ideal composition of CuI.

For the analysis of the marshite from Chuquicamata, it was necessary to eliminate numerous tiny crystals of atacamite. This was done with the electromagnet, and before the analysis was made the sample was estimated to contain less than l/6 atacamite. The amount of chlorine present in the analysis would demand abofi 5/6 of atacamite. In addi- tion, the presenceof such a large quantity of atacamite would require more copper and less iodine than the analysis shows, while in its present form the Cu-I ratio of the analysis agreeswell with the ideal composition of CuI. It therefore seemsbest to consider most of the chlorine of the 632 O. W. JARRELL

Chuquicamata material to be present in the mineral itself, and not due to mechanically admixed atacamite. Upon heating in the closed tube, purple fumes of iodine are given off, which condensehigher up the tube as purple crystals of iodine. Occurrence.AII the specimens come from the south end of the open pit, within 12 to 24 meters of the original surface. The marshite, with atacamite, Iines fractures in the sericitized and kaolinized granodiorite. In this section of the mine, antlerite, otherwise the principal oxidized copper mineral at Chuquicamata, is not stable. The marshite crystals penetrate the atacamite and perch upon it. Probably the deposition of the minerals was essentially contemporane- ous, although some of the atacamite formed after the marshite. The atacamite does not contain any appreciable amount of iodine, for its indices of refraction are normal, and no iodine is given ofi during heating in a closed tube. It is interesting to note that on several occasionsafter blasting opera- tions, small clouds of purplish smoke, which was shown to contain iodine, have been observed at Chuquicamata. Examination of the rock from which this gas came revealed no visible marshite or salesite.Diesel- dorf (1899) has shown that several copper deposits in the arid portion of New South Wales similarly contain small quantities of iodine, al- though it is not present in sufficient quantities for its mineralogical nature to be recognized.

Or-rvBNrrn CugAszOs.Cu(OH)z At rare intervals specimens of olivenite have been found at Chuqui- camata. Excellent crystals were measured by Professor Palache and yielded the forms m(ll}), o(101) and e(011), shown in Fig. 4. The following optical properties were measured on specimens sent by Mr. Zeihen: a:1.780 Biaxial positive p:1.820 2V near90" z:1.865 r

Olivenite, when found here, has always been as small crystals em- bedded in porous aggregatesof a green arsenate, which approximates chenevixite in composition. Occasionally small amounts of an uniden- tified blue arsenate (?) are also present. MARSHITE FROM CHILE 633

In previous descriptions of Chuquicamata, arsenatesare stated to be relatively rare in the oxidized ore. Actually chenevixite is a common mineral in the center of the pit, and is occasionally found elsewhere. Specimensof it and the other arsenates that still contain remnants of enargite show that, where primary enargite is exposed to oxidation, copper arsenateswill form in situ.

Frc. 4. Olivenite.

LrnBruBrqrrn CusPzOs'Cu(OH)z A few specimens of dark olive-green crystals of libethenite perched on white, extremely kaolinized granodiorite were found on the west side of Bench E-4. This immediate area of kaolinized granodiorite contains suftcient copper to be classifi.edas ore, although most of the rock in the vicinity is iron stained and barren of copper. Professor Palache first recognized the identity of the crystals by crystallographic measure- ments. The crystals show a combination of unit prism zz(110) and unit pyramid s(111)about in the proportionsof Fig. 5. The measurementsare very close to those of crystals from the older known localities: No. of Mean measured angles Calculated faces6pQp {110} 3 46"13', 90'00' 46"10t 90"00' {1111 4 46 16 45 24 46 10 45 23 634 O. W. JARRELL

These crystals are apparently the first to be described on which there is no trace of the dome e(011),usually more prominent than the pyramid. The optical properties of the crystals, which agreewell with previously described material are:

a:1.702 Biaxial negative F:I.7as 2V: *80" z:1 .785 />?r, strong

Frc. 5. Libethenite.

With the exception of turquoise, libethenite is the only phosphate so far found in the oxidized ore. rts rare occurrencecan only be regarded as exceptional, and it cannot be fitted into any general paragenetic sequencefor the deposit. DanepsxrrB NaNOs.Na2SO4.H2O Wetzel (1928) briefly noted that darapskite occurred at Chuqui- camata. Actually it is a common mineral in certain zones within a few meters of the surface. The following distinctive optical properties make it easily recognizable in aggregateswith other minerals:

a:1.390 Biaxial negative B:r.481 2Y: +25" :r:1.488 /> ?,strong

The material available proved to be too intimately admixed with other minerals to permit of separating a pure sample for analysis. MARSH]TE FROM CHILE 635

Darapskite is found chiefly in veins up to six inches wide, intimately mixed with kroehnkite and bloedite. Good crystals have not been found. In some casesmirabilite or epsomite are also present in these veins. Both of the latter minerals are crystalline, colorless, and glassy when first broken out from the rock, but even in that dry air they break down to a white powder within a few minutes. Shellac or varnish are of little use in halting this process. The presence,on the hillside approximately 2000 feet above the floor of the Calama Basin, of nitrate and iodine, both characteristic of the nitrate deposits in the basins of the Atacama desert, is of interest in connection with the origin of these deposits. After the sulphate radical, nitrate is the most abundant acid constituent of the sroundwater at Chuquicamata. WurnBNrrn PbMoOa Mr. Zeihen recently recognizedwulfenite in specimensfrom the north end of Bench E-4. It is found in dark gray masses,distinctive because of their adamantine luster, and associated with fine-grained antlerite. The optical properties are:

uniaxial negative e:2.30+0.02 o:2.38+0.02 Wet tests for molybdenum were also obtained. These specimenscame from what has been recognized as an area of primary molybdenite mineralization, in which most of the molybdenum was retained as lindgrenite during oxidation. Minor amounts oI galena have also been recognizedin polished sectionsof the ore.

AcrNowr,BoGMENTS Professor Charles Palache and Dr. Harry Berman have aided the writer in the preparation of this paper. Mr. George Switzer kindly helped in the preparation of the illustrations.

Rnlnnnxcls AnrNorr': Geol.Fiiren. i. StochhotrmFdrh.,44, M+ (1922). Berov: Am. Mi.nerol.,23, 669 (1938). DrnsBr.oonr:Proc. Roy. Soc. N.5.W.,33, 160 (1899). M.qlsn: Proc. Roy.Soc. N 5.W.,26,326 (1893). Prrecnexn Jenner.l:Am. Mineral.,24, 388(I939). Pwon: Minerol.Mag., 13, 189 (1902). SpnNcIrn:Mineral,. Mag., 13,45 (1901). Worznr.: Chetni,ed.er Erde,3, 411 (1928).